Allocating resources in multi-core computing environments

• US 10,020,979 B1
• Filed: 03/25/2014
• Issued: 07/10/2018
• Est. Priority Date: 03/25/2014
• Status: Active Grant

• Current Assignee: A10 Networks Incorporated
• Original Assignee: A10 Networks Incorporated
• Inventors: Oshiba, Dennis, Xiao, Hong
• Primary Examiner(s): Zand, Davoud
• Application Number:
  US14/225,377
• Time in Patent Office: 1,568 Days
• Field of Search: 709203, 709217, 709218, 709219, 709234, 709241, 709243
• CPC Class Codes: H04L 41/00 Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 43/0876 Network utilisation, e.g. volume of load or congestion level

Abstract

Provided are methods and systems for allocating resources in a multi-core computing environment. The method comprises selecting, by one or more processors, at least one dedicated core for execution of a resource allocation algorithm. After selection of the dedicated core, the dedicated core allocates, based on the resource allocation algorithm, a network resource to a client. Furthermore, the dedicated core assigns the network resource to network packets associated with the client for processing by data cores. After the assigning of the network resource, the data cores process the network packets according to the allocated network resource.

Claims

1. A method by a first processor core of a plurality of processor cores for allocating resources in a multi-core computing environment, the method comprising:

assigning a level of service to each client;

receiving a request from a client;

determining, by a dedicated core of the plurality of processor cores, the level of service associated with the client;

executing, by the dedicated core, a resource allocation algorithm, wherein the executing the resource allocation algorithm includes;

collecting statistics associated with processing of a network traffic associated with the client, the statistics including one or more of a volume of a transmitted network traffic associated with the client and a volume of a queued network traffic associated with the client; and

based on the statistics, selecting a quantum of a network resource for processing of the network traffic associated with the client;

allocating, by the dedicated core, using the resource allocation algorithm, the quantum of the network resource to the network traffic associated with the client, the client including at least one client processor, wherein the network resource is allocated to the network traffic associated with the client based on the level of service; and

assigning, by the dedicated core, using the allocating, other processor cores of the plurality of processor cores to processing network packets of the network traffic associated with the client, the other processor cores processing the network packets according to the allocated quantum of the network resource.

2. The method of claim 1, wherein the processing of the network packets includes one or more of sending, receiving, forwarding, consuming, holding, queuing, delaying, and modifying the network packets.

3. The method of claim 1, wherein the assigning of the other processor cores to processing the network packets includes adding a quantum of the network resource to each of the other processor cores.

4. The method of claim 3, wherein the quantum of the network resource is a portion of an available network resource.

5. The method of claim 3, wherein the quantum of the network resource is allocated to each of the other processor cores using a number of network packets queued on the respective other processor core.

6. The method of claim 5, further comprising:

readjusting periodically the quantum of the network resource; and

allocating further the quantum of the network resource, using at least one of the allocated network resource, an amount of an available remaining network resource, an amount of a consumed network resource, a total amount of a network resource waiting to be processed, and an amount of a network resource waiting to be processed by each of the other processor cores.

7. The method of claim 1, wherein the client includes an application or a group of applications.

8. The method of claim 1, wherein the client includes a user or a group of users.

9. The method of claim 1, wherein the client includes a host or a group of hosts.

10. The method of claim 1, wherein the client includes a network or a group of networks, and at least one type of traffic.

11. The method of claim 1, wherein the plurality of processor cores communicate with each other using at least one of message passing and a shared memory.

12. The method of claim 1, wherein the resource allocation algorithm includes a quality of service algorithm guaranteeing the level of service to the client.

13. The method of claim 1, wherein the resource allocation algorithm is configured to limit resource consumption by the client.

14. The method of claim 1, wherein the resource allocation algorithm is operable to implement sharing of the network resources between one or more clients.

15. A system for allocating resources in a multi-core computing environment, the system comprising:

at least one hardware processor configured for;

assigning a level of service to each client; and

selecting a first processor core of a plurality of processor cores as a dedicated core;

the dedicated core configured for;

receiving a request from a client;

determining the level of service associated with the client;

executing a resource allocation algorithm, wherein the executing the resource allocation algorithm includes;

collecting statistics associated with processing of a network traffic associated with the client, the statistics including one or more of a volume of a transmitted network traffic associated with the client and a volume of a queued network traffic associated with the client; and

based on the statistics, selecting a quantum of a network resource for processing of the network traffic associated with the client;

allocating, using the resource allocation algorithm, the quantum of the network resource to the network traffic associated with the client, the client including at least one client processor, wherein the network resource is allocated to the network traffic associated with the client based on the level of service; and

assigning, using the allocating, other processor cores of the plurality of processor cores to processing network packets of the network traffic associated with the client; and

the other processor cores processing the network packets according to the allocated quantum of the network resource.

16. The system of claim 15, wherein the dedicated core allocates the network resource for processing the network packets by adding a quantum of the network resource to each of the other processing cores.

17. The system of claim 15, wherein the quantum of the network resource is a portion of an available network resource.

18. The system of claim 15, wherein the quantum of the network resource is allocated to each of the other processor cores using a number of network packets queued on the respective other processor core.

19. The system of claim 18, wherein the dedicated core periodically readjusts the quantum of the network resource and allocates the quantum of the network resource based on the allocated network resource, an amount of an available remaining network resource, an amount of a consumed network resource, an amount of a total network resource waiting to be processed, and an amount of a network resource waiting to be processed by each of the processor cores.

20. A non-transitory computer-readable storage medium having embodied thereon a program, the program being executable by a first processor core of plurality of processor cores to perform a method for allocating resources in a multi-core computing environment, the method comprising:

assigning a level of service to each client;

receiving a request from a client;

determining, by a dedicated core, the level of service associated with the client;

executing, by the dedicated core, a resource allocation algorithm, the first processor core being selected by a processor as the dedicated core for executing the resource allocation algorithm, wherein the executing the resource allocation algorithm includes;

collecting statistics associated with processing of a network traffic associated with the client, the statistics including one or more of a volume of a transmitted network traffic associated with the client and a volume of a queued network traffic associated with the client; and

based on the statistics, selecting a quantum of a network resource for processing of the network traffic associated with the client;

allocating, by the dedicated core, using the resource allocation algorithm, the quantum of the network resource to the network traffic associated with the client, the client including at least one client processor, wherein the network resource is allocated to the network traffic associated with the client based on the level of service;

assigning, by the dedicated core, using the allocating, other processor cores of the plurality of processor cores to processing network packets of the network traffic associated with the client, the other processor cores processing the network packets according to the allocated quantum of the network resource.